The project will investigate how the human visual system correctly integrates information about different stimulus attributes. Previous work in the field has attempted to isolate the basic stimulus features used by the visual system, but has ignored the problem of how these features are correctly combined in perception. In these experiments, normal adult subjects will be briefly presented with a number of stimuli that can be described in terms of independent attributes, such as shape and color. The main dependent variable will be the errors that subjects make in combining information about these features (i.e. report shapes and colors in incorrect combinations). The research tests the proposal that the visual system uses information about perceptual structure for feature integration. The hypothesis is that features within different perceptual units are constrained by the visual system from combining. This hypothesis will be tested with perceptual units defined in three ways: 1. By Gestalt principles; 2. By subjective organization; 3. By linguistic structure in letter strings. In addition, the research will use errors in feature integration to investigate a number of questions about perceptual structure and information processing within each stimulus domain. For example, the first section of the proposal will investigate the cause of visual hidden-figures effects, and the relation between figure-ground organization and spatial attention. The second section of the proposal investigates the role of experience on perceptual organization. The final section of the proposal uses the feature integration of paradigm to study functional units in work perception such as syllables, consonant clusters, etc. An understanding of how the visual system normally integrates information has relevance in two health related areas. First, visual agnosia can be thought of as an extreme case of a failure to correctly combine stimulus features. Understanding feature integration in normal adults may shed light on this pathology. Second, this research is relevant to the problem of dyslexia and reading acquisition. Dyslexia may be related to a failure to appreciate constraints on the combinations of features into letters, or letters into words.